![]() |
Likely, a mix issue
I have worked with Proseal and other similar sealants often at my day job over the last 35 years, and this does occasionally happen when during the mix, a pocket of the curing paste is not mixed in fully. We now use power mixers which helps, but if you have other areas that are fully cured with no problems, I suspect this may be what happened in your case.
|
I agree with Bill that it was most likely a mixing issue. I have a heated temperature controlled epoxy cabinet in my shop from building a Cozy MKIV, so I kept my Pro-Seal in the cabinet during the tank builds. Kept at around 110* IIRC, the Pro-Seal is very EZ to work with and mixing in the hardener was a breeze.
|
Very common (search the forums), and yes I have seen it many times.
Tank sealant will act just like yours did if it has a prolonged exposure to being constantly whetted with fuel, and a source of oxygen. |
Quote:
|
Does this imply that the tanks must be full after every flight?
Dave |
|
You described exactly my thoughts when it happened to me (goo) #4 post and the factory will tell you the same thing. Use paint thinner,acetone, or any number of bad chemicals with a small wire brush and a plenty of elbow grease. Remove any pro seal that has blue stain not just the gooy part it has been effected also. I learned the hard way had to go through this twice. Mix well new batch of pro seal and apply liberally.
Good luck G.P. |
Quote:
It has nothing to do with how the sealant was mixed. There is no oxygen inside a tank that contains fuel. Fuel vapors are heavier than air.... they displace the oxygen. Different story on the outside of the tank. |
Quote:
The interior is wetted, but not constantly evaporating, like a small weep, so is a weep necessary for the exterior to have deteriorated? Or did this begin from the interior? What are the bench/lab validation results for dipping in fuel, what are the quantified acceptable parameters for long life? Depending on the answer -the question is - will ALL tanks eventually have leaks at the top where the fuel does not touch when parked? |
Quote:
If there is a leak or seep where sealant is always being exposed to evaporating fuel, the sealant will become soft. If you want to know the chemistry explanation you will have to contact the sealant manufacturer, but they will tell you the same thing |
Thanks everybody. It seems like this is a normal phenomenon and maybe not the result of an improper mix ratio. Who knew?
I'm in the middle of cleaning up all the soft stuff and fuel and should have ample amounts of new sealant in place in the interior of the tank in the next day or two. I'll also lop some more on externally just for good measure. Thanks again! |
So I presume the takeaway here is : See a leak? Drop everything and fix the leak?
|
There are a lot of variants of "Proseal". At least 3 different manufacturers and 2 or 3 varieties for each manufacturer. What stuff are you using on the fuel tanks?
|
Quote:
http://standards.sae.org/air4275/ ![]() Quote:
|
Quote:
Since I have seen it numerous times (once a very mild case on my own airplane), and the word that someone at Flame Master acknowledged it can happen, that has always been enough information for me. They way I look at it... I don't need to know why.... I just need to know to always be inspecting for fuel leakage, and then fix it as soon as it is detected. |
Quote:
BTW, oxidation is a maybe, but I soaked a cured sealant sample in 3% hydrogen peroxide nearly 4 years without softening. It did eat the metal can. And I recently examined a softened polysulfide sealant sample taken from inside a tank. It wasn't fully reverted (more like not very sticky week-old chewing gum, or old-fashioned window glass glazing putty), but certainly not normal-as-cured. The owner removed the rivets and easily peeled the skin off the ribs. Samples from the top of the tank (vapor/splash exposure) and the bottom of the tank (submerged exposure) were similar. Samples of the same sealant from outside the tank (zero fuel contact) were solid/normal. Built circa 2005, sealant brand unknown; checking now. Postscript: Sealant was Chem Seal (Flamemaster) 3204-B....same stuff common to RVs. |
Maybe it just a coincidence, but everytime I've seen this it was with the flamemaster stuff.
|
Quote:
|
Quote:
I have some work experience (more than 30 yrs) with PS890, PR1422 and PR1440 and never heard of this problem with avgas. Problems from inproper mixing and curing in an environment that was to cold, yes, but not properly mixed and cured. |
Quote:
![]() |
l'll say it a bit differently. I have not worked withthe Flamemaster product, only the PRC product. Has anyone seen this problem with the PRC product?
|
Quote:
This can begin to tell us the failure rate (or MTBF) in calendar days, etc. |
So is the theory that the leak occurred because the sealant turned to goo.... or the sealant turned to goo because the leak wasn't addressed?
|
Quote:
|
Quote:
|
I've been down this road with quick built tanks that did not leak but showed dozens of blisters at rivets and am not convinced anyone has hit on what is going on with the sealant going goo or blisters that do not show fuel stains or building technique applying the stuff.
What I do know, and it is not conclusive or prove anything for sure, is if you apply enough of the stuff inside the tank at ribs and covers and rivets, the tanks do not leak and the blisters do not appear - at least so far in my history with this stuff. The RV-7A now owned by Jerry Cochran is not showing blisters to my knowledge (about 6 years,new tanks I built) and neither is the RV-8 (about 2 years). So the message is - use lots of sealant, don't be bashful about it, cover every possible place a fuel can leak out with much of it. With any luck at all, it won't leak. And don't get all hung up with the cosmetics of how it looks. The worst cosmetic effect is seeing blue on the outside of the tank. |
Quote:
Returning to sealant reversion.... This from a study conducted for the Navy: ![]() The complete paper is available for download here: http://oai.dtic.mil/oai/oai?verb=get...fier=ADA087267 There are some interesting clues regarding oxidation of fuels (which apparently leads to sealant reversion) in this paper: https://web.anl.gov/PCS/acsfuel/prep...08-90_1277.pdf 100LL also has significant quantities of antioxidant added to meet the requirements of long-term storage. This list of approved antioxidants is from the UK 100LL standard; North America and AUS assumed to be similar: ![]() I would suggest a theory ("theory" = an unproven supposition needing work to determine if it is true or false) that when 100LL gets outside the tank (like one of these fuel leaks) the antioxidant evaporates or is combined/dissolved (i.e. all used up), leaving the leaked fuel subject to a high level of peroxide formation. Acid formation may be involved/required per the paper quoted above. The acid contribution seemed to be an unknown at the time of publication (1980). I am not a chemist, which is why I would like to see a detailed technical explanation from the sealant manufacturer. Failing that, perhaps we can crowd source a qualified individual from our ranks; we have a huge collection of intellect and education lurking on VAF. |
Quote:
|
Soft ProSeal
This thread sparked my interest since I recently observed this on an aircraft that was in service for nearly 10 years. Thing is, the softening was occurring on every sealed joint on both tanks. Even those that showed no staining from leaks. If the softening is the result of the presence of oxygen and gas, well guess what, they are both constantly present. If this is the case proseal is a poor choice.
After looking at the multiple threads on this, I tend to agree with the observation that it appears to be manufacturer specific, especially in occurrences after years in service. I wonder if an informal tally of occurrences would point this out. The difficulty is many aircraft are no longer owned by the builder so identifying the product manufacturer will be difficult. Further complicated by whether or not the builder documented the source of the sealant. I have inquired with some jet service center techs that regularly use proseal type products on fuel tanks and none have seen this. The common response was "the older it is the harder it gets". The opposite of what is occurring in these threads. Of course these shops are going to be very conscientious regarding MIL-Spec and source to meet certified standards. Experimentals are not held to that. Maybe we all need to take stock in the old adage: you get what you pay for! There are simply some corners that should not be cut. |
FWIW - 3M - Response
Since this came back up in another thread http://www.vansairforce.com/communit...d.php?t=142499
I will update my (dead end) probe. I called 3M and eventually talked to a PhD in their engineering/research for the polysulfide (?) products. I sent her the pictures and thoughts from this thread about the possible cause of the color additive and the ASTM for 100LL. There were several discussions and exchanges. I followed up with her in a month (it takes a long time in large companies to fit in projects and get answers). The answer was that the even with 100LL she could not find any evidence that their product variants had exhibited "reversion". She was clearly not a marketing type and got very technical, very quickly, and seemed be genuinely interested in understanding the issue, but since there was no direct link to 3M produced material, the investigation stopped. That seemed reasonable. If someone does use 3M and has this let me know and I can give you the contact information and get it to this engineer/scientist. Meanwhile, my fingers are crossed for tanks. |
To add another layer to the mystery..
Has this happened on any sloshed tanks? My tanks are over 23 years old(pro seal and sloshed) and have been fueled for about two years now, and knock on wood, no leaks. Maybe the extra layer helps.
|
Quote:
|
I've read the threads on the slosh, some do and some don't have problems. If I had to do it over again, I wouldn't slosh, but I did it exactly by the book and no hints of any type of problem so far. Keeping fingers crossed on that.
|
i am wondering what mix ratios people are using? i have been using Proseal, flame master and other types of sealants as well as a lot of epoxy resins for decades on commercial aircraft and helicopters, and have never seen sealant fail as described, unless its mix ratios are not correct, or the resin/sealant is 2 years out of date or longer.
Working with Epoxy resins, we have a general rule, of never divide by more than 1. Usually resins come in tins of 100Grams of part A, and, for example, 22 of part B so, divided by 1, would be 50 grams of part A and 11 Grams of part B. never smaller. using this example, (which is Hysol 9309.3NA structural bonding adhesive) we have 100 % resin, and 22 % part B. 1 gram mistakes in measurements = 1% variation in the mix for the resin, and 5% in part B. so, if using half, 50 resin /11 part B, then 1 gram error will = 2% but a 1 gram error in part B is nearly a 10% variation, and this will effect the quality of the final cure and its properties. so, if you are trying to be frugal, and use 25 grams of resin, then you need 6.5 of part B, and a 1 gram error when mixing will result in a mix that doesn't cure correctly or at all. and this 1 gram error can come about by hardener (part B) that sticks to the mixing stick, or side of the container and never mixes, or unmixed resin in the corners of the container. or just scales that arnt calibrated properly, or have an accuracy variation over differing weight ranges. Im not suggesting that this is the reason for the failed sealant described in the original post, but it is something that has to be considered when mixing sealants for critical areas such as fuel tanks, and pressurisation areas, and especially when mixing structural adhesives for composite work and bonding structural components. |
I used 10:1 by mass, and a digital scale that displayed to .01 grams*. I'd get the hardener to within .02 grams of the target. I figured that was close enough with a claimed accuracy of .02 on the device.
|
I've mixed test samples with radically wrong ratios. They all cured, which is not a huge surprise; the manganese dioxide is a catalyst. There are minor differences in properties (shore, flexibility, tensile), but nothing radical.
The cured samples were then soaked in both a MEK/lacquer thinner mix, and in 3% hydrogen peroxide....for years. Eventually the peroxide ate the cans, but the sealant samples were intact. None reverted to goo. This is an illustration from an Essex Chemical paper on polysulfide sealants. I have no useful knowledge of chemistry at this level, but I'm sure we have readers who do. ![]() |
Quote:
|
GooGate Status Report
Well since this thread got resurrected I might as well give a status report regarding GooGate.
After a performing a cleaning of the area as best as I could, my fix was to lay some bigger fillets of sealant over the seams inside the first bay of the tank. I also cleaned up the goo on the outside really well and reapplied fresh sealant. Here we are more than a year and a half later and no more leaks or goo! :) |
Completely anecdotal, but in the course of working at two different helicopter MRO facilities, none of the old heads would touch a can of Flamemaster sealant if it was provided with a kit, even for non-fuel sealing.
The first time someone told me not to even bother to unwrap the Flamemaster sealant that came with a medical floor kit, I figured it was just a case of "The old man's grumpy again", which was a well-known phenomenon with this particular individual. After the 3rd person at two different facilities said the same thing (almost verbatim), I figured there might be something to it. Again, completely anecdotal. Outside A&P school, I've never worked on anything that runs on avgas, so I don't have first-hand experience with this phenomenon. I've seen sealant soften, wrinkle, and debond in integral fuel tanks (mid-'80s to mid-'90s vintage S-76 fuel tanks - whether it was the cause of the leak, or the result of the leak, is something we never attempted to determine), but never turn to goo. |
| All times are GMT -6. The time now is 08:16 AM. |